Journal of Coastal Research 808--820 Fort Lauderdale, Florida Summer 1995

Dynamic Structures and Their Sedimentation Effects in Huangmaohai Estuary, China

Chaoyu Wut and Shuyao Yuanf

tlnstitute of Coastal and tlnstitute of South China Estuarine Studies Oceanology Zhongshan University Chinese Academy of Science Guangzhou 510275, China Guangzhou , China ABSTRACT _

WU, C. and YUAN, S,,1995. Dynamicstruetures and their sed imsntation effecta in Huangmsohai Estuary Chins. Journal of Coostal R..earch, (1(3) , 8OS-820. Fort Lauderdale (Flor ida) . ISSN 0749·0208. '

Based ,on data c~Uected (rom three cruises of hydrographic surv ey carried out during 19~1991 and num.rI~ modelling the present paper provides a brier analys is on the dynamics of Huangmaohai Estuary, emph88lZlng on the low frequency o( Bow patterns or circulation. Labeled wster parcel trsjectories are caJcu~ated to ~.veal the Lagrangian motio,n. Velocity distribution of the estuary indicates a low snergy zone In the mid estuary that acts 88 a sediment trap. Three large scale dynamic structures are found in tbe estuary along the North and East Channel which have profound effects on the modern sed imentation. They tl1~ (rom north to south (1) vertical de nsity driven circulation, (2) gorge jet current and (3) horbontel CIr~8t10n. The effecte of eetuarine dynamics on the sedimentation process , especially those concerning the river mouth bar, are then discussed.

INTRODUCTION ment of Lagrangian motion is in the same order Huangmaohai Estuary is one of the eight out­ of magnitude as tidal excursion. Lagrangian re­ lets of the Pearl River. It is important for various sidual currents, especially their distribution pat­ economic developments of the west Guangdong, terns in the estuary are shown to be responsible China. Several multi-discipline surveys have been to the transport of suspended sediments. carried out since the 1980's. Because its topog­ The density-difference-driven residual cur­ raphy is complicated with channels, shoals and rents play an important role in Huangmaohai Es­ tidal flats are situated alternatively and the fresh tuary as in many estuaries in the world. However, water discharge is subjected to significant season­ in a particular estuary, the local topographical al variations against the unequal semi-diurnal effects may become the dominant process in a . Its dynamic characteristics are not well un­ certain area . The is of importance in derstood. Modern sedimentation encompasses the the Huangmaohai Estuary and in other broad es­ entire complex of interrelated physical, chemical tuaries in South China where long term circula­ and biological processes that bring about the ac­ tion is a concern. cumulation of sediments. Hydrodynamic analysis STUDY AREA that reveals the basic physical process is funda­ mental to a deeper understanding of the modern Huangmaohai Estuary is located in the west of morphological and sedimentational processes in Pearl River Delta. Tanjiang River and Hutiaomen an estuary. River flow in from the north and are the prin ­ Based on field data and numerical modelling cipal sources of fresh water (Figure 1). Geomor­ techniques, in view of Eulerian and Lagrangian phologically, it is a drowned river valley. It is 40 motion, the present research attempts to identify km long and 1.9 km wide in the upper, north end some of the most significant dynamic structures and has maximum width of 34 km near the mouth . that have profound effects on flow patterns and The estuary has a surface area of 548 km", Two long term transport of water and sediments. The sets of rocky island line up near the mouth in the macro-scale properties of the estuarine dynamics orientation of ENE-WSW. The estuary is bathy­ are the main concern. Huangmaohai Estuary is a metrically complex. A scoured deep channel ex­ convection -dominated system . The net displace- tends 20 km from the head towards south which is mainly the result of fresh run -off and strong 94065 received and accepted in revision 7 July 1994, tidal current in the upper estuary. The North Structures and Sedimentation 809

Channel is about 300-900 m wide and 12 m deep. Two channels are found in the lower estuary. West and East Channel are separated by Damang Is­ land. In the mid estuary, North Channel and the Meters XINHUI channels in the south are separated by a broad • APR 1988 shoal, or river mouth bar where the water depth 22 0 .~~.~; @) JAN is only 2 m which is the major obstacle for navi­ 1989 10' :"~"'i$ 0 JUN 1991 gation. Mean near the head is 1.24 m. . r:-,";;I~,:~:Jj Annual mean fresh water discharge from the two "', , rivers is 1,262 m-/sec, 0 5 TIDAL MODELS KM ~ 0 A two dimensional tidal model is applied to ZHUHAI simulate the fluid dynamics of the estuary. Since the actual transport of water and suspended sed­ ~ iment is a Lagrangian process, a numerical ap­ proach is also incorporated to the 2-D model to r

gu( u2 + V2) O.5 (1) C2H av av av av - + u- + v- + w­ at ax ay az Figure 1. Location map of study area and current meter moor­ ing. = _gar _ fu + kPaWylWyl ay pH

(2) C(x, y) is Chezy coefficient f is parameter ar a a P, Pa are density of sea water and air re­ - + -(Hu) + -(Hv) = 0 (3) at ax ay spectively k is friction coefficient where are speed in x and y directions. u(x, y, t) is depth-averaged velocity in x direc­ Boundary conditions: tion u r = uop(x, y, t) v(x, y, t) is depth-averaged velocity in y direc- V = vop(x, y, t) tion r g is gravitational acceleration where uopand vopare velocities in the open bound­ x is horizontal coordinate (south) ary. For controlled boundary: y is horizontal coordinate (east) I', = rop(x, y, t), H(x, y) is water depth, H = h + r h(x, y) is water depth below plane z = 0 In the present model, observed tidal boundary r(x, y) is water elevation above plane z = 0 was used. For closed boundary,

Journal of Coastal Research, Vol. 11, No.3, 1995 810 Wu and Yuan

HUANGMAOHAI

D 5 KM

HUANGMAO HAl ESTUARY Figure 2. Map of bottom topography. Figure 3. Flow field, flow direction is tangent to the local iso­ bath. (Wn) = 0 where W is velocity vector and n is the direction vector normal to the boundary. The area of in­ yo) at time to' As time elapses, the position of terest is Huangmaohai Estuary. A North South water parcel can be given by oriented finite difference schematization was made X=x,,+l1 with grid size x = y = 600 m. The shallow-water equations (1) to (3) have been solved by an al­ y = Yo + ~ (4) ternative direction implicit (AD!) finite difference ~ method. Time step of 300 seconds gives enough with the initial conditions that 11 = = 0 at time t = to' Thus the Lagrangian velocity (u., v,) can stability. The model is calibrated so that both calculated tide elevation and current velocities be expressed in terms of the Eulerian tidal veloc­ are compared with observed records with fair ity: agreements. Lagrangian Motion When the transport of dissolved and suspended where the position of the water parcel is (x", Yo) matter is of interest, the actual transport is char­ at t = to' The Lagrangian displacement (11 , ~) can acterized as a convection dominated process which be obtained from integrating the differential suggests that the transport is Lagrangian motion. equation of a streakline: The fate of suspended matter should be derived by a Lagrangian mean rather than by an Eulerian (6) mean (CHENG et al., 1986). In order to analyze the Lagrangian motion of labeled water parcel, the Lagrangian velocity components (u., VI) in the or x, y directions are defined as the velocity of a water parcel at time t which is released from (x",

Journal of Coastal Research, Vol. 11, No. 3, 1995 Structures and Sed imentation 811

(A) VELOCITY Cllis TIDAL VELOCITY

(e) nDAL VELOCITY

JIlJ[: 1J HOUR

Figure 4. Velocity isogram in different tidal phases. (D) shows th e residual velocity.

•Journal of Coastal Researc h. Vol. 11, No. 3. 1995 812 Wu and Yuan

~, season. This low energy zone acts as a sediment = fto+T [utx, + 1/, Yo + t), trap for both suspended and bed load from inland and marine sources. vtx, + 1], Yo + ~)] dt (7) Lagrangian Motion where T is one or more tidal periods. Eq. (7) is The verified 2-D numerical model is used as the an integral equation of (1], ~). With appropriate based-line flow field and the movement of marked initial and boundary conditions, Eqs. (1) to (7) water parcels can be calculated from Eq. (7). describe completely the dependent variables (r, Shown in Figure 5 are the labeled water parcel u, v, u., V 1], ~). h trajectories computed in a Lagrangian sense over TIDAL CURRENT ANALYSIS a period of diurnal and semi-diurnal tides. La­ grangian net displacement in Huangmaohai Es­ Eulerian Velocity Distribution and its Effects on tuary is smaller than tidal excursion as required Suspended Sediment Transport but it is large enough to be the same order of An important property of the tidal current ve­ magnitude as tidal excursion. As a comparison, locity in the estuary is that the tidal current, both the Lagrangian net displacement is one order of magnitude and direction, is mainly controlled by magnitude smaller than tidal excursion in a semi­ the . The direction of the current is enclosed estuary with little fresh water (CHENG tangent to the local isobath (Figures 2 and 3). The et al., 1986). This indicates the strong effects of spatial distribution of current velocity based on fresh water run-off on the tidal current and sed­ the three cruises of hydrographic survey and the iment transport. Note that the tidal excursion of verified 2-D model indicate the existence of a low marked water parcels initiating from both upper energy (velocity) zone in the mid-estuary, which and lower estuary is substantially greater than coincides with the river mouth bar, or mid shoal. that from the mid-estuary. Note also that the wa­ Figure 4A-D shows the current velocity in differ­ ter parcels initiate from either upper or lower ent phases in a tidal cycle. Shown in Figure 4D North Channel, they end closely in the mid-es­ is the isogram of Eulerian residual current. The tuary over the river mouth bar. This unique La­ maximum ebb current exceeds 110 em/sec in the grangian motion reveals the effects of flow dy­ North Channel and 90 em/sec in the lower estuary namics on the sedimentation process. The small while it is less than 60 ern/sec in the mid-estuary. Lagrangian current velocity during the turning of During most of the time in a tidal cycle, current tides is favored for sediment deposition. Water velocity over the mid-bar is substantially lower parcels or suspended sediment initiating from the than that in both upper and lower estuary. The lower estuary are carried to deep water by the decrease of current velocity significantly affects strong West Guangdong nearshore drift. Figure 5 the transport rate of suspended sediment and sed­ shows thatthe marked water parcels starting from imentation process. According to BAGNOLD (1966) the upper East Channel do not discharge to sea the suspended load discharge q8 expressed as dry through lower East Channel, instead, after pass­ weight per unit time and width is given by: ing the gorge, turns to the west, joins the West 1'8- I" U 2 Channel and causes serious siltation between Da­ -- q8 = O.OlTo - (8) mang and Hebao island. I" W where 1'8 and I" are specific weights of sediment DYNAMIC STRUCTURES AND SEDIMENTATION and water respectively, To is unit tractive force exerted by the flow on the bed, U is average flow Based on hydrological survey, remote sensing velocity over the , w is settling ve­ images and the numerical model, three large scale locity. When flow carrying sediment from both dynamic structures are found in the estuary along river and the sea slows down in the mid estuary, the North and East Channels which have a pro­ the sediment transport capacity decreases rapidly found effect on the sedimentation. which in turn may cause suspended sediment to accumulate in the water column over the mid shoal. Vertical Density Driven Circulation Satellite picture, hydrographic survey and 2-D Vertical density driven circulation, or non-tidal model indicate that a high sediment concentra­ gravitational circulation in some estuarine liter­ tion zone exists in the mid estuary during flood ature, is the vertical circulation in an average sense

Journal of Coastal Research, Vol. 11, No.3, 1995 Structures and Sedimentation 813 over one or more tidal periods. The circulation is driven by the variation of horizontal pressure along the depth between the baroclinic and barotropic components. The horizontal pressure gradient at "- any depth z in an estuary is ~ N (9) where o;is the density at the sea surface. In strat­ ified conditions, the integral term increases with depth so that the contribution of the surface slope is gradually compensated. Flows induced by the first, density, part of Eq. (9) are called baroclinic flows, and those due to the sea surface gradient are barotropic flows. When the first term exceeds the second term after a certain depth, the water flows up estuary with lighter fresh water flows down to the sea in the upper layer. At the toe of the density circulation near the bottom, the up­ stream flow meets the down-stream flow, which is often called the 'null point' where serious silt­ ation occurs. The density circulation is mainly affected by estuarine bathymetry, river discharge, coastal water body and . The develop of the circulation is most sensitive to fresh water run­ Figure 5. Labelled water parcel trajectories in a tidal cycle off. During the flood season, the river discharge indicate a convection-dominated system. can turn the entire estuary to brackish water. The isohaline of 150/00 is pushed down to the south of Hebao and Daqin Island near the mouth. The gradient is in the same direction as the barotropic 300/00 isohaline can reach Yanan near the head of gradient and enhances it at depth. This tends to estuary in low water season. When the limit of reduce the shear stress in the water column by saltwater moves down in flood season the distance enhancing the landward flow near the bed, and of salt water intrusion increases due to the strong the shear is relatively concentrated near the bot­ stratification. Shown in Figure 6a-c are the ve­ tom. This is shown clearly in all the observed locity profiles in a complete tidal cycle measured velocity profiles. On ebb, the baroclinic pressure in low water season in April, 1988. The up-stream gradient is in the opposite direction as the bar­ flow near the bottom can be found in station 1111 otropic gradient and reduces it at depth. The re­ (Figure 6a) moored near the head of the estuary. sult is substantial shear (Figure 7a). Stratification Figure 6b and c show that the depth (thickness) is then increased and boundary shear stress is occupied by the up-stream flows increases from reduced on ebb relative to the neutral value. The starting at 0.8 height at station I to 0.7 height at ebb-flood asymmetry has important conse­ station 1112, and finally 0.3 height at station 1113. quences for sediment transport (Wu, 1989). Most The up-stream net current velocity near bottom aspects of the ebb-flood asymmetry discussed here is 5-10 em/sec. Shown in Figure 7a and bare have been confirmed in other systems, e.g., the velocity profiles measured in flood season (June, Fraser River Estuary (GEYER, 1988) and the Da­ 1991). The seaward flow controls the entire North wamish (PARTCH and SMITH, 1978). Channel; the toe of the density circulation is sit­ The vertical circulation in the East Channel is uated between stations HI and H2 over the mid strongly disturbed by a jet current system near shoal. In most of the time, Yamen Estuary is par­ the narrow gorge between Sanjiaoshan and Da­ tially mixed. Weak stratification occurs in the high mang Island. To the south of the gorge, a hori­ water season. The ebb-flood asymmetry exists in zontal circulation develops in the Cross-sea where all observations. On flood, the baroclinic pressure the vertical circulation disappears. Shown in Fig-

Journal of Coastal Research, Vol. 11, No.3, 1995 814 Wu and Yuan

5TART TIME:12:00 , APRIL 1988 STN' III 1

~ - 100 - 80 - 60 - 40 - 20 0 liD GO eo 100 120 14U c;'+_-.JL-_--'-__L-_--'-__t-_-'-__...L__L-_-'-__.L-_-'-__...L_-+ '" C), o,

r . r <, e­ o <, , N N '

CD u» r . .I" I- C) CJ ,­ lL' o, W w Oeo ro O C), o,

C)

- 100 ·-00 - GO - 40 - 20 0 20 40 GO BO 100 120 ]110 iso VE LOC ITY PROFILE l e M/ s )

START TIME:12:00, APRIL 19BB STN' III 7

~ - 100 - 00 ·GO -'10 - 20 0 20 40 60 00 IJO 1?rP. 9 --i--- '-- --'-- - -'--- -'-- -!- - --'-- - --'-- - -'--- '-- -'--- --1 '? 2 11 11

, I I n , 1'\ w.J u J 0_ m L ) C), ci, IIJO CJ

- 100 -8 0 - 60 ~ O -~ 0 ~ 40 60 BO 100 120 VELOCITY PR OFILE [ e m/ s l

START TIME: 12:00, APRIL 1988 STN: III 3

~ - 90 - 75 -GO -4 5 -30 -IS 0 IS 30 45 60 75 90 0 c;' + __'--_-'-__--'-__'-_-'-__+__...L__-'-_-.JL-_-'-__--'-_---t :~ );nmt \ llZl:l'!9 0 I"LJWc 15 20 19 N N o, o, r <, 0 N '

CD :!: ": r I- 0 0 1­ n, ' ' lL W W Oro (DO c;' o, IBO --,.. o

-90 15 GO , ', 5 -30 lS 0 IS 30 60 7':, vu.oc i TI PRm-1LE. le Mh I Figure 6. Velocity profiles measured in Apr il 1988 at station (a) 1II1 (b) 1II2 and (e) 1II3. The gross line represents the tidally mean velocity.

Journal of Coastal Research, Vol. II, No.3, 1995 Structures and Sedimentation 815

START TIME:11:00, 6.29-30,1991 STN: H'

~ -120 -100 -80 -60 -40 -20 0 20 liD 60 80 100 120 140 160 18~ 9~_...... J..__...l...-_---l-_----L__....l...-_""'-+-__.L....-_-'--_----I..__J....-,_...... L.-_----L__....l...-_----'-__+-9 10 23 1122 8 1 m Iti tdII 13 20 19 Iii 18 17 15 16 N o o I I

:r::---- ::r. ::r. ----::r:: <, 0 o <, N ' IN :r:: CD CD• ::r:: ~ 0 o ~ 0- 1 lO­ W W Dco roO o 9 J SER ~ o o

-120 -100 -80 -60 -40 -20 0 20 40 00 80 100 120 140 160 180 VEL~CITI PROFILE (cm/sJ

11 :00 6.29-30, 1991 HUANGMAO SEA STN' H2 21 59'19"N 113 06'40"[

120 -100 uu bU 40 20 o 20 40 60 80 100 120 16LP 9-l------L--l----L------..!..----L----+---~-....l...------L--J.....----l..------:.---l------t-9 \7 \e., \ 5 I o / I ) ~. I o <, 1 N

ill ::rill • ::r:::: I- 0 o I­ (LI In.- W w Oro roO a o J J SER ~ o o

-120 -100 -80 -60 -40 -20 0 20 40 60 80 100 120 140 160 VELOCITI PROFILE (cm/sl Figure 7. Velocity profiles measured in June, 1991 at station (a) HI and (b) H2. The gross line represents the tidally mean velocity.

ure 8 are velocity structures in the lower estuary. Jet Current They show a complicated combination of surface The narrow scoured trough in the upper East and bottom Ekman flows, density flow and gra­ Channel between Damang and Sanjiaoshan Is­ dient flow which differ from the structure of the land (Figure 1) has played an important role in density dominated flow in the upper and mid es­ the sedimentation process. The gorge is approx­ tuary. The existence and variations of the density imately 1,500 m wide. The water depth exceeds 5 driven circulation have strong effects on sedi­ m in the trough and decreases to 2-3 m to both mentation. In low water season, mixing is strong sides of the gorge which separates the mid shoal and the 'null point' is moving back and forth with­ in the East Channel into two parts. The seaward in the North Channel. Sediment supply from both and landward changes in the water depth are con­ the river and sea is low and erosion takes place sidered to be the response to an efflux system or, in the river mouth bar (Wu, 1991). Severe siltation simply, the deceleration flows. Tidal current ve­ occurs in flood season when the river carries a locity exceeds 1.0 m/sec in the upper layer and great amount of sediment to the estuary and at 0.8 m/sec in the lower layer in station 1113 near the mean time when the 'null point' that serves the gorge and decreases rapidly in both directions. as a sediment trap is right over the mid shoal. Shown in Figure 9 are the gradual changes of tidal

Journal of Coastal Research, Vol. 11, No.3, 1995 816 Wu and Yuan

STN: H.o

10 20 30 40 50

(\J (\J o o I J

~ ~ . I o <, 1 N

to to I . . I l- 0 o I­ 0- 1 lO- W W Oro roD o 9 I SE.R~

-60 -SO -40 -30 -20 -10 0 10 20 30 40 so 60 VELOCITi PROFILE (cm/sJ

START TIME:9:00, 6.27-28, 1991 STN HO

c: -60 -50 40 3U -20 -10 10 20 ]U l10 :JU 50 c: 0 0 I I

0 0 I ,

~ ~ ~ ~ :r . . I <, 0 o <, I ~ I ~

I~ to. :r l- 0 o I- 0- J I 0- W W Oro roD 0 0 1 I

0 0 -; -; -60 -SO -40 -30 -20 -10 0 10 20 30 40 SO 60 VELOCIT'( PROFILE (cm/sJ Figure 8. Velocity profiles measured in April 1991 at station HO, (a) north component and (b) east component. The gross line is the tidally mean velocity.

current velocities calculated from the 2-D model layers respectively. Observations taken near the which decreases with the increase of the distance gorge indicate that the values of F, range from 2­ from the gorge. The stability of the interface be­ 8, the maximum reaches 8-40 during the low wa­ tween the upper fresh water layer and salt water ter season in the winter. F, ranges from 0.4 to 2.0 below depends on the value of the densimetric during the flood season when buoyancy increases. Froude number F, given by The distance from the top of the shoal in the north to the gorge is approximately 8 km and this dis­ F.=~ (10) tance is 4 km in the south. Studies point out that 1 y:ygh the distance from the top of a depositional shoal to the jet mouth is about 4-6 times the width of where U is the velocity of the upper layer, g is the the jet. This shows reasonably close agreement acceleration of gravity, h is the flow depth of the with our observation. Based on observations (sta­ upper layer and l' the density ratio tion II 3 in January, 1989, station V4 in July, 1989) taken near the gorge, the Eulerian residual cur­ (11) rent is directed upstream in all depths. The efflux where PI and Puare the density of lower and upper system has restrained and actually replaced the

Journal of Coastal Research, Vol. 11, No.3, 1995 Structures and Sedimentation 817

POINT#' VELOCITY ALONG THE GORGE POINT#2 FROM 20 MODEL. POINT#3 POINT#4 0 POINT#5 (\J

0 0

0 co 0

0 CD en <, 0 L:

0 0 w :::jf w 0..... 0 en 0 (\J 0

0 0 0 0 2 6 8 10 12 IlJ 16 18 20 22 24 26 TI ME (HLJURJ Figure 9. Jet current velocity variation. Point #1 to #5 line up along the East Channel from north to south. Point #3 is located in the middle of the gorge, the distance between each calculated point is approximately 800 m.

density driven circulation near the gorge as a Island. One branch flows north passing through dominant dynamic process. The gorge efflux sys­ the Damang Gorge; the main stream turns west tem is sedimentologically significant. The depo­ and joins the West Channel. The mean velocity sition of the shoal to the north of the gorge is the in the Cross-sea ranges 5-15 em/sec and reaches result of density circulation acting as a sediment 20 ern/sec outside the East Outlet. Kelvin wave, trap and the deceleration of the jet flows as well. topography, and the West Guangdong drift all The shoal to the south of the gorge is mainly the have their contribution to the development of the result of the efflux system. Since the jet system circulation. The mechanism of this circulation is is physically stable, the shoals in both sides of the not yet completely clear. Kelvin wave is the bal­ gorge have not shown significant changes since at ance between and the horizontal least the 1930's based on the navigation charts gradient when tidal waves enter an estuary. Water published from 1861 to the present. As a com­ elevation and velocity under the Kelvin wave are parison, the shoal in the West Channel has ex­ given by: perienced significant seaward migration in the same period. The inner slope of the shoal has been r= Roef/Ccos(ai - ~x) scoured and the isobath of 5 m migrated 40 m per year seaward in the last 40 years (YANG, 1992), at the mean time the outer slope advanced 36 m/yr U = ~ Roef/cCOS (fJt - ~ x) (12) to the sea (Figure 10).

where c is the phase celerity of the tidal wave, (J Horizontal Circulation is angular frequency of the tidal waves, f the Cori­ Observations and the 2-D model reveal a hor­ olis parameter and R, the amplitude of the tidal izontal circulation in the Cross-sea in the lower wave. Under the Kelvin wave, both tidal range estuary. It is a residual circulation. Water flows and current velocity in the eastern part are greater in from East Outlet between Gaolan and Hebao than that in the western part. A counterclockwise

Journal of Coastal Research, Vol. 11, No.3, 1995 818 Wu and Yuan

0 o 2 0 < Z - o W ui 0 I /:;~~:==-:~~-:-.:.:~.------5 ...... , . (A) <, ~ ::> <, o -, I :r: t- N '-- fu o 4: --- 0 10 t- .-En

horizontal circulation is then prone to develop. marine side mainly comes from Modaomen and Yuedong nearshore drift flows from ENE to WSW Jitimen, also outlets of the Pearl River located to along the coast. It can be considered as a geo­ the east of Huangmaohai Estuary. Based on the strophic flow which is the result of the dynamic measurement of navigation charts from 1937 to balance between the Coriolis force and the water 1977, the average depositional rate of the bay is elevation gradient with the high water along the 1.28 em/yr. There are three zones with high de­ coast. When the drift reaches the estuary, the positional rate: the mid-shoal, the western mar­ gradient no longer exists and the drift flows in ginal shoal and eastern marginal shoal. Since large­ the estuary through the East Outlet. The residual scale reclamation projects have been carried out circulation has certain effects on the water and in both the eastern and western marginal shoals, sediment transport in the lower estuary. The up­ the maximum natural depositional rate occurred stream residual flow restrains the sediment en­ in the mid-shoal which reaches 4.0 em/yr. Nega­ tering the sea through East Outlet. Since the shelf tive depositional rate occurred in both the North water with low levels of suspended sediment flows Channel and the Lower estuary. The high depo­ in, it is favored for the East Outlet to maintain a sitional zone in the mid-estuary is coincident in deep channel free from serious siltation. position with the low energy zone. In October 1990 a testing navigation channel NATURAL DEPOSITIONAL RATE AND was opened in the mid-shoal connecting the North CHANNEL REDEPOSITION Channel and the East Channel. The total length The main source of suspended sediment in the of the artificial channel is 7.5 Km, the bottom estuary is from the fresh water which counts for width is 60 m, the bottom elevation is 4.0 m, and 77 % of the total suspended sediment input to the average dredged depth is 1.14 m. estuary (YANG, 1992). Sediment input from the was taken in a regular base to monitor the rede-

Journal of Coastal Research, Vol. 11, No.3, 1995 Structures and Sedimentation 819

position. Figure 12 shows the echo sounding lon­ gitudinal profile of the channel. The survey in­ dicates that severe deposition occurred during river floods and storm surges. Slightscouring took place in low water season in winter. The rede­ positing rate is more than 60 em/yr. It is no sur­ prise that the artificial channel was completely refilled in less than two hydrographic years during which no maintenance dredging occurred. The low energy zone and null point deposition mechanism have profound effects on the sedimentation.

CONCLUSIONS (1) Observations and the 2-D numerical model reveal a low energy (velocity) zone in the mid estuary that acts as a sediment trap for suspended and bed load from both inland and marine sources. (2) Labeled water parcel trajectories calculated over a complete tidal cycle reveal some of the important features of the Lagrangian motion. The net Lagrangian displacement is in the same order as tidal excursion, which indicates the seaward flow is a dominant process in Huangmaohai Es­ tuary. (3) Three large scale dynamic structures which have substantial effects on sedimentation are Figure II. lsogram of depositional rate in the estuary, data found in the estuary along the North and East collected from 1938 to 1977 (modified from VAN(;, 1992).

SH 1991 .08 NAVIGATION CHANNEL SOUNDING CH 199 1. 10 CH 1991.08 H U A I ~ G M AO SEA 1990.1 2- 1991.1 0 0 CH 1991.06 If) CH 1990.12 N I · . . 0 . . _-- ·- .---- - 0 · .. M ·. . I .: " ' , __oro' - -- , -- -t - - -:-- - --;-- -- , -- . ;... . .:.- ---:•.. -;. •. - ; .. ••:. •. - -:•. -- ;--- 0 · . . _- ~- . . -: .- ._~_ ._-~ -- - ~ -- - . : . _- - --: . . _ . ~-- _ .: ~ .. --:- . . . ~ .. - If) · . . ·.·. . . M _ __ 0 - • • •• _ • •• •• •• . ••• • _ .__ _ _ ••• I · . .. . ·...... 0 . . . . 0 i' I 0 If) I I- cr (L I . , r: \ . .;... W 0 0 .....: ..:... . : , :-.. . ~. - _~ "" _.-:.'\ ...... / I \. "/ ..... 0 , ,- • \ . .j .:. .. . ~./ : tD I : . · . ...· ...... I "i":" .:.. .. ;.. .: ..,. . \;'I ~>- I .. .. . -.·- - -"\. · ·- - . · · -· r -· · ·.-· · · ..····.···· ,···-.-· ··..···-,···

0 ~ _ - - -_ ". ------• • -- _ . _. . " • , .-- . I .. __ ,- . " _ ._ --_ -_ -, _ -.. If) ... . , . . . . . ·...... ··. . . . . u-i · . " I +-----'--,---'-----i--'-----,-----'-- r--'---.- -'- -i----''-II·- - -,--- .,------,---,----,---r---.----' 0 7 14 2\ 20 35 42 49 56 63 70 77 84 91 98 o[S TANCE ('" 1OOM) o is the h ea d o f d reded c h o n n el Figure 12. Repeat echo sounding of the testing channel profiles. SH-shouJder of the channel. CH-channel bottom.

Journal of Coastal Resea rch. Vol. II, No.3, 1995 820 Wu and Yuan

Channels. (a) Density driven circulation migrates lipse. In: Physics of Shallow Estuaries and Bays, New from the head of estuary in dry season to mid York: Springer-Verlag, pp. 102-113. estuary where it causes serious siltation in flood GEYER, W.R., 1988. The advance of a salt wedge front: Observations and dynamical model. In: DRONKERS, J. season. (b) Near the gorge between Damang and and VAN LEUSEEN. Physical Processes in Estuaries. Sanjiaoshan Island, the efflux system becomes a New York: Springer-Verlag, pp. 81-195. dominant dynamic feature in the East Channel. JAY,D.A., 1988. Residual circulation in shallow estuary: This system is physically stable; the shoals in both Shear, stratification and transport processes. In: CHENG, R.T. (ed.), Residual Current and Long-Term sides of the gorge have not shown significant Transport, New York: Springer-Verlag, pp. 49-63. changes since at least the 1930's. (c) The hori­ JOBSON, H.E., 1981. Temperature and solute-transport zontal circulation in the lower estuary is consid­ simulation in streamflow using a Lagrangian reference ered the result of the combination effects of Kel­ frame. U.S. Geological Survey Water Resources In­ vestigation 81-2, 165p. vin wave, topography and West Guangdong drift. PARTCH, E. and SMITH, J.D., 1978. Time-dependent The upstream residual flow restrains the sedi­ mixing in a salt wedge estuary. Estuarine Coastal ment entering the sea through East Outlet and Marine Research, 6, 3-19. directs the flow to the pass between Damang and PRITCHARD, D.W., 1956. The dynamic structure of a Hebao Island. coastal plain estuary. Journal of Marine Research, 15,33-42. Wu, C.Y., 1988. The effects of settling velocity on the LITERATURE CITED formation of high concentration zone in a well-mixed estuary. Tropical Geomorphology, 9(1), 63-78. BAGNOLD, R.A., 1966. An approach to the sediment Wu, C.Y., 1991. Hydrodynamics of the Huangmaohai problem from general physics. U.S. Geological Sur­ Estuary, Pearl River. Technique Report, Institute of vey, Professional Paper 422-1, 37p. Coastal & Estuarine Studies, Zhongshan University. CHENG, R.T. and CASULLI, V., 1982. On Lagrangian re­ (in Chinese) sidual current with applications on South San Fran­ YANG, G.R., 1992. Sedimentation in Huangmaohai, Pearl cisco Bay, California. Water Resource Research, 18(6), River Estuary. Technique Report, Institute of Coastal 1652-1662. & Estuarine Studies, Zhongshan University. (in Chi­ CHENG, R.T., et al., 1986. On Lagrangian residual el- nese)

Journal of Coastal Research, Vol. 11, No.3, 1995